Ontology-Based Type Mapping for Open-Domain Question Answering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current information retrieval systems lack the capability to accurately answer factual questions from a collection of documents, whether structured, unstructured, or both, due to challenges in understanding queries, finding relevant documents, and extracting correct answers.
Innovation Solution
A computing infrastructure and methodology that uses a candidate answer scoring technique to map lexical types to query types, leveraging ontologies for type matching and axioms for considering type inclusion/exclusion, to generate answers to questions by determining the disjointness between types and scoring candidate answers based on these relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional information retrieval methods are used to answer questions from unstructured or structured documents, then the system can retrieve documents containing relevant information, but the system fails to accurately extract and deliver correct factual answers
Solution Approach 1:
The patent introduces an intermediary component (answer extraction module with verification mechanisms) that bridges the gap between document retrieval and accurate answer delivery. This intermediary processes retrieved documents to extract factual answers while verifying their correctness against multiple sources or criteria, thereby improving answer accuracy without requiring complete system redesign
Solution Approach 2:
The patent replaces traditional keyword-based mechanical search mechanisms with semantic understanding approaches that comprehend the meaning and context of queries. This substitution enables the system to accurately identify and extract factual answers from unstructured and structured documents by understanding semantic relationships rather than relying solely on keyword matching
2Measurement precision
If the system attempts to understand and process natural language queries to find relevant documents and extract answers, then answer quality improves, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing and indexing documents with semantic metadata before query processing. This preliminary structuring of information enables faster retrieval and processing during actual query execution, as the system can quickly locate relevant pre-processed content without performing complete semantic analysis from scratch
Solution Approach 2:
The patent segments the question answering process into distinct modular stages: query understanding, document retrieval, answer extraction, and verification. This segmentation allows each stage to be optimized independently and processed in parallel where possible, reducing overall processing time while maintaining answer quality
3Quantity of substance
If the system relies on general ontologies and world knowledge for open-domain questions, then more data is available for answer extraction, but the lack of domain-specific knowledge reduces answer accuracy
Solution Approach 1:
The patent implements a multi-functional knowledge base architecture that integrates both general ontologies and domain-specific knowledge sources. This universal system can adapt to different question types and domains by selecting appropriate knowledge sources, thereby maintaining both broad data availability and domain-specific accuracy simultaneously
Data Source
AI summary
A computer-implemented system, method and program product generates answers to questions in an input query text string. The method includes determining, by a programmed processor unit, a lexical answer type (LAT) string associated with an input query; automatically obtaining a candidate answer string to the input query from a data corpus; mapping the query LAT string to a first type string in a structured resource; mapping the candidate answer string to a second type string in the structured resource; and determining if the first type string and the second type string are disjointed; and scoring the candidate answer string based on the determination of the types being disjointed wherein the structured resource includes a semantic database providing ontological content.